Since the first-ever LV phase I clinical trial against AIDS in 2001, 38 phase I-II and two phase II-III trials exploiting HIV-based LV as gene delivery vehicles have undergone authorities' scrutiny; three of them are still under review. The largest number of trials comprises monogenic disorders, some of which with large incidence such as Cooley's anemia β-thalassemia major (4 trials). Cancer and infectious diseases, mostly HIV-1 infection, follow. Commonly, the number of patients enrolled in phase I/II trials is limited, but not that in phase III. Thus stable packaging cell lines for 2nd (LTR-based) and 3rd (SIN-based) LV generation are urgently needed to cope LV large scale production demand for phase III trials hopefully attainable in the future in a larger number. LV production grounded on transient protocols is indeed impractical for global application under a safety, cost and reproducibility standpoint.
An increasing body of evidence indicates that LV, the most recently developed viral integrating vectors for gene therapy, are broadly applicable to transduce either terminally differentiated or cycling cells, ideal to sustain long-term transgene expression and safer than what was initially feared. The experience accumulated on Moloney murine leukemia virus (MoMLV) gamma-retroviral vectors (γRV) over the last two decades guided the fast progress on LV delivery system, whose development originated by the necessity of overcoming the inability of retrovirus to transduce non diving cells. In particular, the generation of self-inactivating (SIN) transfer vectors makes the prospect of a large use of LV in human clinical trials more feasible1 provided the expansion and optimization of an as much efficient manufacturing process. However, in contrast to γRV, which can be produced by several human and murine commercially available packaging cell lines, LV are currently produced not only for research-grade, but also for GMP-grade, almost exclusively by transient transfection. This technology is expensive, difficult to standardize and scale-up and requires numerous downstream validation tests. Furthermore, the risk of replication competent lentiviruses (RCL), possibly arising through recombination between viral sequences in the packaging and transfer vector constructs, is a rare, but more likely event during transient than stable production.
The development of a retroviral-equivalent stable packaging cell line for LV turned out to be slower and more difficult because, as opposite to gamma retrovirus, the expression of lentiviral proteins, such as the env, protease, and some accessory proteins is toxic for human cells. To overcome this problem the accessory genes, present in the very early versions of packaging cells, were later removed in the latest generations. First generation SIV- and HIV-based LV packaging cell lines were obtained from either monkey Vero, or human COS, HeLa and 293 adherent cells2-5, engineered with lentivirus genomes carrying few crucial modifications such as the removal of the packaging signal. The gp120 env and most accessory genes were in fact maintained. The resulting LV titer was very low2-5, and more importantly the possible application of these vectors was necessarily restricted to CD4+ T cells for anti-AIDS gene therapy approaches. Later, gp120 env was substituted with the glycoprotein derived from the vesicular stomatitis virus (VSV-G) and all accessory genes were removed because proven dispensable for an efficient LV production. To prevent the toxicity also described for VSV-G, its expression was conditionally induced by a variety of different systems, such as the Tet, ecdysone, Rev and the combination of Tet and cumate6. Similarly, to reduce the toxic effect of the viral protease during clone selection, the conditional expression of the gag-pol gene by the Tet and the combination of doxycycline and cumate drugs have been described6. In all these systems gag-pol, rev and env genes were integrated by transient transfection of plasmid DNA, followed by drug selection and cell cloning.
One of the crucial issue for the implementation of a truly stable packaging cell line is the choice of the best viral gene delivery vehicles. Most researchers integrated the gag-pol, rev and env genes by transient transfection of plasmid DNA, followed by drug selection and cell cloning2-9. This technology is known to suffer over time from gene silencing and gene loss10, which can both jeopardize the long-term stability of the packaging clone.
Alternative gene delivery vehicles have been disclosed particularly in STAR11 and in the more recently developed GPRG-TL-2012 packaging cell lines where the gag, pol, and rev genes were integrated into HEK293T cells by MLV-shuttle vectors. Two copies of the recoded gag-pol gene were stably integrated in STAR, whereas no such information is available for GPRG-TL-2012. As opposite to STAR, where the env gene were transfected, in GPRG-TL-20 all the remaining viral genes were introduced by SIN-MLV.
Several systems exist that allow stable integration of foreign genome into host cells. Palombo et al., 199813 disclose an hybrid baculovirus-AAV vector for specific integration into host cells. Such vector appears to be very effective if it includes rep gene in the same hybrid baculovirus-AVV vector. There is no mention in this reference of the construct of the present invention let alone the suggestion of using this kind of system for LV production.
Over the last almost two decades, several attempts to generate stable LV packaging cell lines have been made. Despite the different technology disclosed, as of today none of these packaging cell lines is employed in clinical trials or corners the market yet. Therefore there is a need of new systems for large scale production of LV that are effective in terms of production capability and are safe, cost effective and reproducible.